The sensitivity and specificity of the routine kidney biopsy immunofluorescence panel is inferior to diagnosing renal immunoglobulin-derived amyloidosis by mass spectrometry.

2019 
Immunoglobulin light chain (AL) amyloidosis is the most frequent type of renal amyloidosis in the United States, accounting for 81% of cases. Accurate typing is crucial for early diagnosis and treatment of immunoglobulin-derived (AIg) amyloidosis and to avoid treating other amyloidoses with potentially toxic chemotherapy. Immunofluorescence (IF) is the first step to type renal AIg amyloidosis but the performance characteristics of this method are largely unknown. Here, we establish the sensitivity and specificity of IF for diagnosing AIg amyloidosis in patients whose amyloid typing was performed by the current gold standard of laser microdissection/mass spectrometry (LMD/MS). Renal biopsy pathology reports originating from several institutions with a diagnosis of amyloidosis and which had amyloid typing by LMD/MS performed at our center were reviewed. Reported IF staining for kappa or lambda of 2+ or more, with weak or no staining for the other AL was considered positive for AL amyloidosis by immunofluorescence. Based on LMD/MS results, of the 170 cases reviewed, 104 cases were typed as AIg amyloidosis and 66 were typed as non-AIg amyloidosis. IF sensitivity for diagnosing AIg amyloidosis was 84.6%. The remaining 16 cases could not be diagnosed by IF due to reported weak staining for all antigens or reported lack of preferential staining for one antigen. IF specificity was 92.4%. Five cases, all amyloid A (AA) amyloidosis, were misdiagnosed as AIg amyloidosis by immunofluorescence. IF failed to accurately differentiate AIg from non-AIg amyloidosis in 12.3% of cases of renal amyloidosis. Relying on IF alone for determining AIg vs. non-AIg amyloidosis may lead to misdiagnosis. Thus, IF has inferior sensitivity and specificity compared with LMD/MS in the typing of AIg amyloidosis.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    29
    References
    11
    Citations
    NaN
    KQI
    []